Sorry if this is a duplicate idea I searched and couldn't believe it hasn't been posted.
This idea is to add Linux RT deployment support for desktop PCs. NI in the past has had a process for purchasing an Phar Lap RT license, then deploying it to a normal x86 based PC. Here are the System Requirements, and a manual.
This opened up the door for desktop PCs to become embedded PCs, controlling NI hardware, but allowing for an upgrade path where embedded cDAQ devices, and even PXI controllers don't have many options.
At the moment there are unofficial ways to get Linux RT on an x86 based desktop with limited success, but no support. These options are outlines in the comments of an idea to allow for a VM of the Linux RT environment. NI could go with a similar route to Phar Lap and sell a license with a list of supported hardware.
Based on the discussion in this thread I would like to propose the following:
My suggestion is to open support for execution of python nodes on Labview Real-Time targets. This is currently not supported by LabVIEW 2018 Current Gen, even though this version supports executing a python node using the new "Connectivity -> Python"-support.
This should be possible to do on the targets using NI Linux Realtime as the operating system, since Linux treats python as a native application language.
I understand that the execution of Python scripts are not deterministic, which will prevent an implementation in time critical code. That I can follow, but it should be possible to use the python node in lower priority threads or non real-time code, for communication with a REST-API, downloading/uploading files, connectivity to online services and so on.
Therefore I propose to simplify this implementation and open the support for Python3.
Some additional implementations are essential for the success:
Selection of Python runtime should be fixed to python3, since python 2 will be without further support or active development from 2020-01-01 going forward.
It would gain the best momentum if it was possible to deploy python from the Device Software installation as part of the build specification of the real-time application.
Python developers can specify their external requirements for additional code by defining a "requirements.txt" or "pipfile.lock" depending on the virtualization method and choice of development style. If the project manager or build specification could read/parse this for the starting point of the application to prevent manual labour and possible errors.
Backwards Compatibility:
Support for this should be from LabVIEW 2018 and forward, since these have support for the python nodes.
The new web interface offered when you install a system image to an RT target is very rudimentary, it offers almost none of the configuration options that we had in the Silverlight-based version. You cannot edit the network card settings, you cannot change the date and time etc. All it does is refer you to use NI Max, which kind of defeats the point of having access to basic config without the need for anything but the web browser.
Are there any plans of an update to the web interface any time soon?
When creating an image of an RT target e.g. we currently rely on it to be able to store a temporary copy of the image prior to transfering it to the host PC. This can often cause the operation to fail (especially now that system images are way bigger than before) because the target does not have sufficient storage. If you have an application/config that is large too and you want that included in the image this requirement can mess thing up too.
This could have been avoided by streaming the image data to the host instead, with little or no intermediate storage.
I noticed there seem to be no way to guarantee the state of an output module controlled by a scan engine in case the RT Application (or the Host Application, depending who is controlling the chassis) crashs. With FPGA one can program some kind of watchdog setting back the output values of a module in case the RT Exe fails. With Scan there just seem to be no possibility.
This is why I think adding a FailSafe Value for a Scan I/O node could be a creat idea. in ase the RT application got aborted or stops without cleanup, the output value would not be random no more but set back to their FailSafe value. I imagine it could look like that:
I can currently write code that uses the SysConfig Restart VI to run some cleanup operations and then shutdown a cRIO.
However, if I rebuild my startup application in LabVIEW, and then want to deploy it (or want to restart from MAX) then I have no easy way to call this code when restarting - the cRIO restarts and any handles are left open/half-open until the other end clears them up.
Could we have the ability to register a shutdown VI that is called whenever the target is programmatically restarted (e.g. by LabVIEW's Project Explorer or MAX)?
I understand that in the case of failure and a hard reset via the Reset button then probably this is impossible/implausible, but in the "normal" workflow, this would be a useful addition.
64bit has been the dominant architecture for a decade; any computer with more than ~2.5GB of RAM must use it after all. It is inevitable that 32-bit machines will cease to be made - maybe not tomorrow, but let's be realistic. Let's get ahead of the times and convert modules to support 64 bit support. Please!
Currently, non default map constants (like the one shown below) are detected as fatal insanities when deploying to Real-Time targets and crash the LabVIEW development environment.
There is a workaround, which is to build the map from other data structure constants. The map constant above was built from the code below. However, it takes longer to execute, and it takes more code. That is okay in some cases, but it would be better if there was an option to simply use the map constant above.
The real-time controllers have a "Time server" IP input in their setup. That is great, but it is not great that the time server has to be a piece of NI software (Logos). If it was possible to specify an NTP server for example (and/or other standard protocols), this feature would be much more useful.
Most of the time we need the PACs to be synced with a third party system.
NI has introduced malleable VI function since LabVIEW 2017. It is a great function to have in advanced LabVIEW application development.
But "Stall Data Flow.vim" function is missing in Real-Time Target OS from timing function palette since 2017.
It is not like you can not use malleable VI in RT side, I have copied same VI from desktop VI to RT VI. It works fine without any error or broken arrow.
NI should include this types of VIs in RT package by default.
RT has Type Specification Structure too, which creates question for me on this missing VI.
If you have any reason(s) to share on this why it is not included on RT side, Please share your thoughts.
During LabVIEW Development was a PXI Real-Time System connected (LAN) next to my Laptop and process Build/Deployment was no problem.
Now the RT has been shipped far away to customer site and i have no longer access to that network.
So I need an other way to distribute and
here is my IDEA #1: (below is another IDEA #2)
FACT : Most of all customers have no knowledge about "How to find the Real-Time System and update the Application on it".
So the new item will built an "Offline Installer.exe" and will run on any customer PC to a full automated update job automatically.
Features:
- create a detailed installation report which can be stored and send back to the developer.
- In case of the Real-Time System isn't connected to any LAN, the customer could use the "Offline Installer.exe" to prepare an USB-Stick which has to be plugged into the Real-Time System. Then a reboot will launch the full automated application update process. (the "RT LEDs" will give feedback when the process is finished)
I tried also another way but without success:
Step 1: Add a ZIP File into the Project Explorer : "My Zip File"
Step 2 : Configure the "My Zip File" => Add "My Real-Time Application"
Step 3 : Build "My Zip File"...but the Result was bad :
But anyway, if the would be successful there are a lot of more steps needed:
- sending the ZIP to the customer
- unpack ZIP
- find Real-Time System at customers network
- a FTP tool is needed to update the system.
- There is no way to stop the running Application (from a PC without NI-Software)
Conclusion:
The update with "ZIP" is far away from a comfortable solution.
IDEA #2
If the Real-Time System has a running webserver (the thing which is using Silverlight),
then the application could be updated by using a webbrowser.
So, first we have to update the webserver on the Real-Time System (as difficult as updating the application)
There are 2 LabVIEW libraries available to create HDF5 files, the problem is, there is no version of the HDF5 library available for NI Linux RT systems, so everyone that wants to use this, and doesn't have enough experience to build the HDF5 library on their own can't use this.
A recent forum post discussed possibilities for "Securing a cRIO": Securing CompactRIO, and identified that LabVIEW project access allows the project to upload new rtexes without a password.
You are then prompted for a password to reset the cRIO, but this can be somewhat similarly achieved with a screwdriver and the reset button... (of course, this assumes physical access to the cRIO, whereas project access does not - but I suspect in the forum users' case, the two were approximately equivalent).
Would it be possible to move the login prompt before the new rtexe and accompanying files are uploaded, rather than for the restart?
Many measurement and process control application run at relatively slow rates (<100Hz). Using SCAN Engine on the CompacRIO for data acquisition is ideal for these applications because you don't need to program the FPGA and all the measurement and control logic can be implemented on the Real-Time controller.
In many cases you want to process your data before you analize it. Currently you only have the ability to get the raw measurement data from the AI modules, so you need to add the data processing code to your existing LabVIEW program. It would be helpful if the SCAN engine could offload some of the data processing (ex. lowpass filter or sample average) to the FPGA and provide the user with already processed data. For example, this functionality can be added to the module configuration page:
When running a VI in development mode while targeting an RT device, you must "Save All" prior to deployment. This is annoying, especially when using SCC. I'm sure that SourceOnly will minimize this effect in LV2010, but the concept still remains: I don't want to be forced to Save All when I don't want my edits (or automatic linking edits) to persist.
Dual network interfaces is often part of the requirements for redundancy, however in such cases it is also very common to specify that the behaviour of borth of these should be identical. You see it in subsea control systems where they have an "A" and "B" channel, you see it topside where the device might need to be on two networks etc.
Unfortunately this is not the case for any of the dual port RT targets from NI. The secondary port is really a second class NIC. It has limited configuration options. It does not support DHCP, you cannot specify a gateway for it - and the code to do programmatic changes to its configuration is not easily available.
Please make the two ports fully interchangable. Port 1 or 2? It should not really matter which one you use.
